Manufacturing condition output apparatus, quality management system, and storage medium

1. A manufacturing condition output apparatus which includes a processor and outputs a manufacturing condition of a product,

the processor being configured to:

output change degree information which is information regarding degrees of change of values regarding defect probabilities corresponding to respective values of a plurality of variables relating to manufacturing of the product from model information of a model generated through machine learning on a basis of manufacturing data of the product and inspection result data of the product, as the manufacturing condition;

display, based on the change degree information, the degree of the change for at least one variable of the plurality of variables on a display screen; and

set, for the at least one variable, the manufacturing condition that is values or ranges of variables where the product becomes a good-quality product or a defective product in accordance with the defect probabilities.

2. The manufacturing condition output apparatus according to claim 1 ,

wherein the model is a non-linear model.

3. The manufacturing condition output apparatus according to claim 2 ,

wherein the non-linear model is a model learned through ensemble learning such as random forest using a plurality of decision trees.

4. The manufacturing condition output apparatus according to claim 1 ,

wherein the processor outputs display data of the display screen for setting the manufacturing condition in association with a probability that the product becomes the good-quality product or the defective product.

5. The manufacturing condition output apparatus according to claim 1 ,

wherein the processor is configured to:

extract variable names of the plurality of variables as important items in accordance with degrees of importance calculated from the model information; and

output information of the extracted important items.

6. The manufacturing condition output apparatus according to claim 1 ,

wherein the product is a die-cast product manufactured by a die-casting machine.

7. A quality management system comprising a data acquisition apparatus, a model learning apparatus, and an estimation result output apparatus,

the data acquisition apparatus comprising:

a first processor configured to:

acquire manufacturing data from a manufacturing apparatus which manufactures a product; and

acquire inspection result data of the product manufactured by the manufacturing apparatus,

the model learning apparatus comprising:

a second processor;

a manufacturing data storage apparatus configured to store the manufacturing data acquired from the first processor; and

an inspection result data storage apparatus configured to store the inspection result data acquired from the first processor,

the second processor being configured to generate a model which estimates whether the product manufactured at the manufacturing apparatus is a good-quality product or a defective product through machine learning on a basis of the stored manufacturing data and the stored inspection result data, and

the estimation result output apparatus comprising:

a third processor configured to:

estimate whether the product is the good-quality product or the defective product on a basis of the acquired manufacturing data using the model and output an estimation result;

output change degree information which is information regarding degrees of change of values regarding defect probabilities corresponding to respective values of plurality of variables relating to a manufacturing condition of the product from model information of the model, as the manufacturing condition;

display, based on the change degree information, the degree of change for at least one variable of the plurality of variables on a display screen; and

set, for the at least one variable, the manufacturing condition that is values or ranges of variables where the product becomes a good-quality product or a defective product in accordance with the defect probabilities.

8. A non-transitory computer-readable storage medium that stores a program that is executed on a computer, the program causing the computer to execute:

processing of outputting change degree information which is information regarding degrees of change of values regarding defect probabilities corresponding to respective values of a plurality of variables relating to manufacturing of a product from model information of a model generated through machine learning on a basis of manufacturing data of the product and inspection result data of the product, as a manufacturing condition;

processing of displaying, based on the change degree information, the degree of changes for at least one variable of the plurality of variables on a display screen; and

processing of setting, for the at least one variable, the manufacturing condition that is values or ranges of variable where the product becomes a good-quality product or a defective product in accordance with the defect probabilities.